一种基于高空间分辨率遥感影像的道路提取方法研究  

A New Approach of Extracting Road from High Spatial Resolution Remote Sensing Images

作　　者：王艳梅[1] 杨学志[2] 

机构地区：[1]安徽三联学院交通工程学院,安徽合肥230601 [2]合肥工业大学计算机与信息学院,安徽合肥230009

出　　处：《昆明学院学报》2017年第3期111-114,共4页Journal of Kunming University

基　　金：安徽省2016年高校优秀青年人才支持计划重点项目(gxyq ZD2016401)

摘　　要：高空间分辨率遥感影像中蕴含着丰富的道路信息,是基础地理信息更新很好的来源.以合肥市城区为研究区域,采用Quick Bird高空间分辨率遥感影像为数据源,开展城市道路信息提取研究.首先,对Quick Bird数据的多光谱波段进行人工采样,获取道路信息的光谱特征.其次,以光谱特征作为判据搜索研究区域上满足道路特征的道路点,得到道路的大致轮廓.最后,运用数学形态学工具对提取的道路结果进行进一步处理.结果表明,该方法可较完整的提取出城区道路网信息.High spatial resolution remote sensing image containing large road information＇s a good source for basic geographic informa-tion updates. Taking Hefei city as the research area, QuickBird high spatial resolution remote sensing image was used as the data source to extract and study the urban road information. Firstly, the spectral features of road information were obtained by manual sampling of multi spectral bands of QuickBird data. Then, the road points were searched by using the spectral feature as the criterion to get the rough outline of the road. Finally,the mathematical morphology tools were used to extract road results for further processing. The results show that the method can be more complete to extract the city road net information.

关 键 词：高空间分辨率 遥感影像 光谱特征 道路 提取 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]